Layer-Wise De-Training and Re-Training for ConvS2S Machine Translation
نویسندگان
چکیده
منابع مشابه
Layer-wise training of deep generative models
When using deep, multi-layered architectures to build generative models of data, it is difficult to train all layers at once. We propose a layer-wise training procedure admitting a performance guarantee compared to the global optimum. It is based on an optimistic proxy of future performance, the best latent marginal. We interpret autoencoders in this setting as generative models, by showing tha...
متن کاملGreedy Layer-Wise Training of Deep Networks
Complexity theory of circuits strongly suggests that deep architectures can be much more efficient (sometimes exponentially) than shallow architectures, in terms of computational elements required to represent some functions. Deep multi-layer neural networks have many levels of non-linearities allowing them to compactly represent highly non-linear and highly-varying functions. However, until re...
متن کاملMinimum Risk Training for Neural Machine Translation
We propose minimum risk training for end-to-end neural machine translation. Unlike conventional maximum likelihood estimation, minimum risk training is capable of optimizing model parameters directly with respect to evaluation metrics. Experiments on Chinese-English and EnglishFrench translation show that our approach achieves significant improvements over maximum likelihood estimation on a sta...
متن کاملCo-training for Statistical Machine Translation
I propose a novel co-training method for statistical machine translation. As co-training requires multiple learners trained on views of the data which are disjoint and sufficient for the labeling task, I use multiple source documents as views on translation. Co-training for statistical machine translation is therefore a type of multi-source translation. Unlike previous mutli-source methods, it ...
متن کاملMaximum Correlation Training for Machine Translation Evaluation
We propose three new features for MT evaluation: source-sentence constrained n-gram precision, source-sentence reordering metrics, and discriminative unigram precision, as well as a method of learning linear feature weights to directly maximize correlation with human judgments. Our source-sentence constrained n-gram precision achieves, among all the testing metrics including BLEU, NIST, ROUGE, ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: ACM Transactions on Asian and Low-Resource Language Information Processing
سال: 2020
ISSN: 2375-4699,2375-4702
DOI: 10.1145/3358414